Evaluation of restoration effectiveness: community response to the removal of alien plants.

Plant invasions are a key cause of biodiversity loss and motivate many restoration programs worldwide. We assessed restoration success of an invaded forest in the Azores using two complementary experimental designs: a before-after control-impact (BACI) design compared a restored and a control (unmanipulated) site over three years, while a control-impact (CI) design evaluated the short-term effects of restoration on restored-control replicated pairs. In both designs, a food web approach was used to evaluate both structural and functional aspects of the restoration. Two years after removing alien plants from the BACI design, there were increases in the abundance of native seeds (110%), herbivorous insects (85%), insect parasitoids (5%), and birds (7%) in the experimental plot compared to the unmanipulated plot. In the CI design, five experimental plots were weeded and paired with five adjacent unmanipulated plots. Immediately following the removal of alien plants within the experimental plots, there was a significant decrease in native plant species, likely attributed to the effect of disturbance. Nevertheless, the production of native seeds increased by 35% in year 1, and seed production of the focal endemic plant, Ilex perado (holly), increased 159% in year 2. Weeding increased the survivorship and growth of seedlings transplanted into the plots, particularly those of alien species. Both experiments provide evidence of the positive effects of weeding cascading through the food web from native plants to herbivorous insects, insect parasitoids, and birds. Two aspects that could prove critical to the outcome of restoration programs deserve further attention: most bird-dispersed seeds were alien, and weeding favored alien over native seedling growth.